1. shopping
2. resources
3. contests and freebies

bs_bs_banner
Environmental Microbiology (2015)
doi:10.1111/1462-2920.12852
Development, diet and dynamism: longitudinal and
cross-sectional predictors of gut microbial
communities in wild baboons
Tiantian Ren,1† Laura E. Grieneisen,2†
Susan C. Alberts,3,4 Elizabeth A. Archie2,3* and
Martin Wu1**
1
Department of Biology, University of Virginia,
Charlottesville, VA 22904, USA.
2
Department of Biological Sciences, University of Notre
Dame, Notre Dame, IN 46617, USA.
3
Institute of Primate Research, National Museums of
Kenya, Nairobi, Kenya.
4
Department of Biology, Duke University, Durham, NC
27708, USA.
Summary
Gut bacterial communities play essential roles in host
biology, but to date we lack information on the forces
that shape gut microbiota between hosts and over
time in natural populations. Understanding these
forces in wild primates provides a valuable comparative context that enriches scientific perspectives on
human gut microbiota. To this end, we tested predictors of gut microbial composition in a well-studied
population of wild baboons. Using cross-sectional
and longitudinal samples collected over 13 years, we
found that baboons harbour gut microbiota typical of
other omnivorous primates, albeit with an especially
high abundance of Bifidobacterium. Similar to previous work in humans and other primates, we found
strong effects of both developmental transitions and
diet on gut microbial composition. Strikingly, baboon
gut microbiota appeared to be highly dynamic such
that samples collected from the same individual only
a few days apart were as different from each other as
samples collected over 10 years apart. Despite the
dynamic nature of baboon gut microbiota, we identified a set of core taxa that is common among primates, supporting the hypothesis that microbiota
codiversify with their host species. Our analysis iden-
Received 28 July, 2014; revised 18 March, 2015; accepted 18
March, 2015. For correspondence. *E-mail [email protected]; Tel.
(574) 631 0178; Fax (574) 631 8149. **E-mail [email protected];
Tel. (434) 924 4518; Fax (434) 982 5626. †These authors contributed
equally to the manuscript.
© 2015 Society for Applied Microbiology and John Wiley & Sons Ltd
tified two tentative enterotypes in adult baboons that
differ from those of humans and chimpanzees.
Introduction
Vertebrate gut microbiota play important roles in host
biology, including immune regulation, energy acquisition,
vitamin synthesis and disease risk (Turnbaugh et al.,
2006; Hooper et al., 2012; Bengmark, 2013; Morgan
et al., 2013). There is mounting evidence that variation in
the functions of gut microbiota is mediated by variation in
gut microbial composition both within and between hosts
(Turnbaugh et al., 2009; Greenblum et al., 2011; Hooper
et al., 2012; Bengmark, 2013; Iida et al., 2013; Karlsson
et al., 2013; Koeth et al., 2013; Markle et al., 2013; Viaud
et al., 2013). However, to date, most such evidence
comes from research on humans and captive animal
models, leaving large gaps in our understanding of the
forces that shape gut microbial composition in wild vertebrates, both between individuals and within the same
individual over time. Filling these gaps is especially important for wild primates, in part because such information
helps reveal which of the forces that shape human gut
microbiota are common across primates, and which are
unique to humans and perhaps a consequence of modern
human lifestyles (Yildirim et al., 2010; Degnan et al.,
2012; Amato, 2013; Moeller et al., 2014).
To help address these gaps, we used cross-sectional
and longitudinal sampling to characterize distal gut
bacterial communities over an unusually long, 13 year
time span in a well-studied population of wild baboons.
Baboons provide an especially relevant comparative
system for understanding variation in human gut
microbiota because of their relatively close evolutionary
relationship to humans, and because baboons lead a
terrestrial, savannah-dwelling lifestyle that is thought to
resemble the ecology of early humans (DeVore and
Washburn, 1963; Codron et al., 2008; Sponheimer et al.,
2013). Specifically, we worked with the Amboseli Baboon
Research Project (ABRP) in Kenya (Alberts and Altmann,
2012), where longitudinal, individual-based research on
demography, developmental milestones, social relationships, diet and climate provided an especially rich context
within which to understand individual-level variation in the
2
T. Ren et al.
Table 1. Sample size information including the number of individuals
and fecal samples used in analyses of the dataset rarefied to 1500
reads.
Individual
Sex
Number
of fecal
samples
BEAM
DUNLIN
OKOT
LEBANON
OCEAN
VIXEN
DRONGO
ECHO
VANGA
GOLON
LAWYER
OXYGEN
HONEY
AMIGO
CABANA
CEDAR
DYNAMO
HEKO
LADHA
LARK
LAZA
PLATO
VOGUE
VORTEX
M
F
M
M
M
F
F
F
M
M
M
F
F
M
F
M
M
F
F
F
F
M
F
F
10
10
10
8
8
8
7
7
7
6
6
6
3
1
1
1
1
1
1
1
1
1
1
1
Range of years
samples were
collected
Age range or age
at time of sample
collection (years)
1994–2001
1996–1999
1996–1998
1997–2009
1997–2000
1994–1998
1996–2009
1995–2001
1995–2001
1996–1999
2001–2001
2000–2001
1999–2000
1998
1999
1998
1998
1997
1998
1997
1998
1998
1998
1997
5.95–13.16
0.72–3.69
1.32–2.81
0.17–12.16
0.6–3.78
17.06–20.84
6.99–19.88
3.43–9.39
2.54–9.31
17.76–20.63
1.12–1.98
1.05–2.43
1.85–3.13
15.07
0.53
2.86
0.94
14.27
4.57
9.71
6.9
8.6
1.08
10.32
structure of gut microbial communities. Our main objectives were to: (i) characterize the basic structure of gut
microbiota in wild baboons, (ii) gain a multivariate understanding of the relative importance of development, social
relationships, diet and climate in predicting gut microbial
community structure, (iii) understand patterns of longitudinal change in baboon gut microbial communities and
(iv) test whether gut microbiota in wild baboons contain a
core set of microbial taxa and enterotypes. Throughout,
we discuss our results in the context of what is known
about human and other primate gut microbiota.
Results and discussion
General patterns in the baboon gut microbial profile
We analysed distal gut microbial composition using 107
fecal samples from 24 baboons collected between 1994
and 2009 (Table 1). For 13 baboons, we analysed multiple
samples (range = 3–10 samples per baboon; time span
between longitudinal samples ranged from 2 days to 13
years). From these 107 samples, we generated 358 428
high-quality 16S rRNA reads, yielding an average of 3350
reads per sample and 7201 total operational taxonomic
units (OTUs) using a 97% identity cut-off. This dataset
was rarefied to 1500 (n = 107 samples from 24 baboons;
Table 1) and 3000 reads per sample (n = 54 samples
from 17 baboons; Table S1). The 1500- and 3000-read
datasets were highly similar in microbial composition at
the OTU level (Pearson correlation coefficient = 0.983,
P = 0.001). Furthermore, Mantel tests correlating compositional dissimilarity in samples rarefied to 1500 versus
3000 reads revealed a high level of congruence between
these datasets (Mantel tests for Bray–Curtis dissimilarities: r = 0.996, P = 0.001; unweighted UniFrac: r = 0.978,
P = 0.001; weighted UniFrac: r = 0.996, P = 0.001). As a
result, the results we present in the main text rely on the
1500-read dataset; where appropriate, we repeat analyses using the 3000-read dataset and present them as
Supporting Information.
Taxonomic assignment revealed representatives from
11 bacterial phyla and 90 bacterial genera. Similar to other
mammalian gut microbial communities, the four most
common phyla included Firmicutes (48.8% of reads),
Actinobacteria (17.2%), Bacteroidetes (7.2%) and
Proteobacteria (4.1%). However, compared with humans
and other primates (Ley et al., 2008; Yildirim et al.,
2010), samples from wild baboons harboured a much
higher percentage of Actinobacteria, of which 97.8%
were assigned to the genus Bifidobacterium, and a relatively smaller proportion of Bacteroidetes (Fig. 1).
Bifidobacterium is dominant in the gut flora of breastfed
human infants (Turroni et al., 2012), where it is thought to
play a role in the digestion of the complex carbohydrates
in human milk. Accordingly, in humans, the percentage of
Bifidobacterium decreases dramatically with age and it
comprises only 3–6% of the adult gut flora. In baboons,
grasses and other fibre-rich foods were common in the
diet, and Bifidobacterium spp. may be important in digesting the high fibre content of these foods.
Notably, nearly one-fifth (20.8%) of reads were unclassified and potentially novel at the phylum level. These
unclassified OTUs were unlikely to be sequencing artefacts because the vast majority (86.1%) appeared in more
than one sample, and their distribution among samples
was indistinguishable from that of classified OTUs (Supporting Information Fig. S1). Moreover, 54.0% of these
unclassified OTUs were ≥ 90% identical to OTUs found in
other mammalian fecal samples (Ley et al., 2008).
Previous studies have found that mammalian gut microbial composition is strongly associated with host diet and
phylogeny (Ley et al., 2008; Yildirim et al., 2010; Hong
et al., 2011; Degnan et al., 2012; Bolnick et al., 2014;
Delsuc et al., 2014). Therefore, we compared the gut
microbiota of our baboon samples to those of other
mammals. As expected, the 13 of 14 fecal samples (one
from each adult baboon) from our study clustered with
other primates, especially those with omnivorous diets
(Fig. 2). Samples from within the same order or diet group
were significantly more similar than samples from different
orders or diet groups (Supporting Information Table S2).
Hence, host phylogeny and diet both seem to play domi-
© 2015 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology
Gut microbiota in wild baboons
3
Fig. 1. Phylum level bacterial composition across 107 samples from 24 individual baboons. Each column represents one fecal sample. Y-axis
values represent the relative abundance of each phylum classified by RDP classifier. Samples are sorted by the relative abundance of
Actinobacteria in the sample.
nant roles in determining variation in gut microbial composition between host species. We note that these
patterns were based on the most abundant bacteria in
each sample because of the small number of reads in the
mammal dataset (Ley et al., 2008).
Juvenile baboons exhibited lower bacterial alpha
diversity, but higher variance than adults
Alpha diversity is an important component of microbial
diversity in the gut, especially in the context of microbiota
development and pathogen resistance (Dillon et al., 2005;
McKenna et al., 2008; Degnan et al., 2012; Flores et al.,
2012; Yatsunenko et al., 2012; Ahn et al., 2013). Furthermore, given that some of our samples were collected
more than 15 years prior to analysis (Table 1), we were
concerned that DNA degradation might affect microbial
alpha diversity in our samples, and hence our ability to
characterize gut microbial composition. However, we
found no evidence that older samples exhibited lower
alpha diversity than younger samples (linear mixed
models with sample age in years as a fixed effect and
host identity as a random effect: species richness:
β = −1.37, P = 0.76; Shannon’s H: β = 0.004, P = 0.89;
Chao1: β = −4.14, P = 0.63; Faith’s phylogenetic diversity:
β = −0.19, P = 0.28).
In addition to sample age, we tested several other predictors of gut microbial alpha diversity, including host age,
sex, rainfall in the 30 days prior to sample collection,
current social group, natal social group, current social
group size, host diet composition, host diet alpha diversity,
adult social rank, and for adult females only, reproductive
state (as pregnant, lactating or ovarian cycling). Because
host identity significantly predicted variation in alpha
diversity for two of the four measures (ANOVA;
species richness: F(12,83) = 2.03, P = 0.031; Shannon’s H:
F(12,83) = 2.78, P = 0.003; Chao1: F(12,83) = 1.39, P = 0.187;
Faith’s phylogenetic diversity: F(12,83) = 1.72, P = 0.078),
individual identity was included as a random effect in all
linear mixed models. In prior studies on humans and
chimpanzees, age was a primary predictor of gut microbial alpha diversity (Degnan et al., 2012; Yatsunenko
et al., 2012). However, in our dataset, neither age nor any
other fixed effects predicted any of the four measures of
alpha diversity in linear mixed models. However, when we
divided samples into juveniles and adults rather than
testing age as a continuous variable, we found that adults
had greater species richness (Wilcoxon rank-sum test;
© 2015 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology
4
T. Ren et al.
Fig. 2. PCoA analysis of the weighted
UniFrac dissimilarities comparing gut
microbiota of baboons to other mammals.
Each point corresponds to a sample coloured
by (A) host taxonomy and (B) host diet type.
Baboon samples are circled in red; the 14
baboon samples were drawn at random
representing one each from the 14 adult
individuals (AMIGO, BEAM, DRONGO,
ECHO, GOLON, HEKO, LADHA, LAZA,
LARK, LEBANON, PLATO, VANGA, VIXEN,
VORTEX) included in our dataset. The
percentage of the variation explained by the
plotted principal coordinates is indicated on
the axes.
W = 1723.5, P = 0.049) and Shannon’s H (Wilcoxon ranksum test; W = 1783, P = 0.019) than juveniles. Additionally, similar to one prior study in humans (Yatsunenko
et al., 2012), we found that infants and juveniles exhibited
significantly higher variance in Shannon’s H than adult
baboons (Fig. 3; Brown–Forsythe test; F(1, 104.876) = 6.988,
P = 0.009). Taken together, these results indicate that gut
microbiota may be less diverse and less stable in young
baboons as compared with adults, suggesting that the
transition to adulthood marks a developmental milestone
in the microbiota.
Variation in microbial composition was best explained
by host age, diet and rainfall
We tested whether several host traits and environmental
factors were associated with variation in baboon gut
microbial composition, including host identity, host age,
host sex, rainfall in the 30 days prior to sample collection,
current social group, natal social group and group size. To
Fig. 3. The relationship between baboon age (in years) and gut
microbiota alpha diversity as measured by OTU Shannon’s H.
Infant and juvenile baboons had higher variance in Shannon’s H
than adults (Brown–Forsyth test F(1, 104.876) = 6.988, P = 0.009).
© 2015 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology
None
Rainfall (P = 0.02)
© 2015 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology
22
38
60
21
Infant
Juvenile
Infant/juvenile
Adult male with rank
information
Adult female with
rank information
24
76
Subset with diet
information
Age, rainfall, sex, individual ID
Age, rainfall, sex, individual ID
Age, rainfall, sex, suckle status, individual ID
Age, rainfall, adult male rank, natal social
group, individual ID
Age, rainfall, adult female rank, reproductive
status, individual ID
Rainfall (P = 0.14)
None
None
None
None
None
Rainfall (P = 0.03), age (P = 0.05),
diet Shannon’s H (P = 0.02) or
diet PC1 (P = 0.02)
Sex (P = 0.03)
None
None
Rainfall (P = 0.03)
Rainfall (P = 0.02), age (P = 0.04),
diet Shannon’s H (P = 0.05) or
diet PC1 (P = 0.02)
Sex (P = 0.04)
None
None
Rainfall (P = 0.05)
None
Rainfall (P = 0.02), age (P = 0.02)
Rainfall (P = 0.02), age (P = 0.01)
107
Main dataset
Age, rainfall, sex, individual ID, social group,
natal social group, group size
Age, rainfall, sex, diet diversity (richness,
Shannon’s H or PCoA axis), individual ID
Genus level
Significant factors at
Phylum level
Dataset
Factors we attempted to include in the model
Number of
samples
Table 2. CCA analysis of environment and host traits that predicted variation in gut microbial community composition rarefied to a level of 1500 reads.
test these factors, we first performed an exploratory principal coordinates analysis (PCoA) on weighted UniFrac
dissimilarities. Overall, 37.5% of the global variation
was explained by the first three principal coordinates
(PC1 = 21%, PC2 = 9.3%, PC3 = 7.2%). Visual inspection
revealed no obvious clustering patterns by any of our
predictor variables (Fig. S2A–E).
To further test which factors best explained variation in
baboon gut microbial composition, we carried out canonical correspondence analysis (CCA) (Palmer, 1993) at the
phylum, genus and OTU levels (Table 2). For the main
dataset (n = 107 samples), we again tested the effects of
host identity, host age, sex, rainfall and aspects of social
group membership. Interestingly, no factors explained
significant variation at the OTU level, perhaps because
closely related species are often ecologically interchangeable (Harvey and Pagel, 1991), and ecological patterns in
bacterial communities may be more apparent at higher
taxonomic levels. In support, some recent studies have
found ecological coherence among higher bacterial taxonomic ranks (Fierer et al., 2007; Lozupone and Knight,
2007; von Mering et al., 2007; Fulthorpe et al., 2008;
Pointing et al., 2009; Philippot et al., 2010; Koeppel and
Wu, 2012). Indeed, we were able to explain significant
shifts in high-level taxa associated with changes in environment. Specifically, rainfall and age predicted significant
variation at both the phylum level (Table 2: rainfall,
P = 0.02; age, P = 0.02) and the genus level (rainfall,
P = 0.01; age, P = 0.02). Interestingly, while age predicted
beta diversity, when baboon infants, juveniles and adults
were considered separately, age was no longer significant, suggesting that either the subsets of samples do not
have enough statistical power or developmental transitions are more important than age per se (Table 2). Host
sex was significant when we considered infants alone,
perhaps due to differences in maternal care as a function
of infant sex (Nguyen et al., 2012). These sex differences
seem to disappear in adulthood, however.
While some prior studies have found evidence for social
group membership on gut microbial composition (Degnan
et al., 2012; Yatsunenko et al., 2012), including in our
own population (Tung et al., in press), the wide temporal
distribution of samples in our dataset probably made it
difficult to detect such effects. We found no other physiological or social effects on gut microbial composition,
including female reproductive state, social group size, or
male or female dominance rank.
The effects of rainfall on microbiota may be linked to
seasonal changes in either diet or drinking water availability. To test the specific effects of diet, we conducted
a second CCA using only the subset of 76 individuals
(excluding infants) for which we had data on the time
spent foraging on different food types. In this new model,
we found several effects of diet (Table 2; Table S4). First,
OTU level
Gut microbiota in wild baboons
5
6
T. Ren et al.
Table 3. Best-supported generalized linear mixed models (Poisson link) explaining variation in abundance of the four most common bacteria phyla
for the main dataset (n = 107 samples). Host identity is modelled as a random effect.
Bacteria phylum
Fixed effects
Estimate
SE
Z
P-value
Actinobacteria
Age
Rainfall
Age
Age
Rainfall
Age
−0.024
−0.003
−0.027
0.008
0.003
−0.0299
0.004
0.0001
0.005
0.002
<0.0001
0.008
−6.599
−17.897
−4.918
4.44
37.21
−3.69
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
Bacteroidetes
Firmicutes
Proteobacteria
diet alpha diversity (Shannon’s H for diet components)
explained significant variation at both the phylum
(P = 0.02) and genus levels (P = 0.05), while diet richness
(the number of distinct food types; Table S3) did not, suggesting that dietary evenness rather than a high number
of dietary components is important to the gut microbial
composition. This pattern runs counter to that seen in
Bolnick et al. (2014), which found that dietary richness
rather than evenness predicted gut microbial diversity in
fish. Second, the dietary tradeoff between the proportion
of time spent consuming grass versus fruit in the diet (diet
PC1; see Experimental Procedures; Fig. S3) was significantly associated with microbial composition at both
phylum level (P = 0.02) and genus level (P = 0.02).
To assess the influence of sequencing depth on our
results, we repeated the CCA using the smaller dataset
rarefied to 3000 reads (n = 54 samples; diet information
on n = 38 samples). We obtained best models similar to
those of 1500-read dataset, although none of the factors
were significant, probably as a result of a loss of statistical
power (Table S4).
Finally, to identify which of the four most common bacterial phyla were associated with differences in host
age, rainfall and diet, we performed generalized linear
mixed models with a Poisson link and host identity as a
random effect. We found that samples collected in rainier
periods harboured a higher proportion of Firmicutes, but
less Actinobacteria than samples from drier months
(Table 3). In terms of host age (measured as a continuous
variable), younger animals harboured relatively more
Actinobacteria, Bacteroidetes and Proteobacteria but less
Firmicutes than older animals, perhaps due to differences
in milk consumption or disease susceptibility in animals of
different ages. For the subset of samples with diet information, gut microbiota from groups that consumed relatively more fruits and less grass harboured higher levels
of Actinobacteria and Proteobacteria and lower levels of
Firmicutes and Bacteroidetes than groups consuming low
fruits (diet PC1, Table S5). Furthermore, the addition of
rainfall significantly improved models with diet factors,
indicating that the effects of rainfall are not solely driven
by seasonal changes in diet. During the dry season, the
baboons drink from small, highly concentrated and qualitatively dirty water holes whereas during rainy months
they obtain most of their water from seemingly cleaner,
transient rain puddles, which may have consequences for
gut microbiota.
Longitudinal sampling reveals that baboon gut
microbiota are highly dynamic
Prior research on humans and chimpanzees has found
that individuals contain distinct gut microbiota, and that
samples from the same individual, even those collected
over a year apart, are more similar to each other than they
are to samples collected from different hosts over the same
time period (Turnbaugh et al., 2009; Caporaso et al., 2011;
Degnan et al., 2012; David et al., 2014). However, we
found no evidence for such effects in our study subjects.
For instance, in the CCA analyses described above, we
never observed a significant effect of individual identity at
any taxonomic level. Similarly, samples from the same
individual were as different from each other as they were
from samples collected from different individuals in the
same developmental stage (mean ± SE weighted UniFrac
dissimilarity: between samples from the same individual = 0.342 ± 0.008; between samples from different
individuals at the same stage = 0.345 ± 0.002, P = 0.35).
However, we note that the sequence depth in our study is
lower than those of the previous studies.
This high degree of dynamism in baboon gut microbiota
can be visualized by plotting pairwise weighted UniFrac
dissimilarities between samples of the same individual as
a function of time between sampling points (Fig. 4). Microbial communities sampled from the same individual a few
days apart were almost as different from each other as
samples collected several years apart. Only 1 of 13 individuals with >3 samples displayed a significant relationship between sampling time interval and microbiota
dissimilarity (Table S6). It is unclear why baboon gut
microbiota appeared to be so dynamic. One possible
explanation is that seasonal variation in the baboons’
diets selects for different gut microbial compositions at
different times of year, as the availability of fruits, seeds
and vegetation fluctuates with seasonal patterns in plant
reproduction. However, such seasonal variation is unlikely
to explain turnover on the scale of days or weeks during
which baboon diets are more consistent. Another expla-
© 2015 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology
Gut microbiota in wild baboons
7
Fig. 4. A time-decay plot of gut microbiota dissimilarity. Each dot represents a comparison between two samples of the same baboon
collected at different time points, with different marker colours representing different baboons. X-axis represents the time span (in days)
between the sample collection times. Y-axis represents the weighted UniFrac dissimilarity. (A) 5000 days. (B) 365 days. The dotted line in
panel A marks 365 days. Correlation between sampling time span and microbiota weighted UniFrac dissimilarity for each individual is
summarized in Table S6.
nation is that wild baboons live in microbially heterogeneous environments, regularly walking through fecal
deposits of other species, drinking from waterholes that
contain fecal material from livestock and wild mammals,
and pulling plants from the ground with their mouths. This
could lead to higher turnover in gut microbial species.
Core gut microbiota
The high degree of inter- and intra-individual variations in
baboon microbiota raises the question of whether baboon
gut microbiota contain a set of core microbial taxa, as is
observed in humans (Tap et al., 2009; Martínez et al.,
© 2015 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology
8
T. Ren et al.
Fig. 5. PCoA visualization of baboon enterotypes (ellipses) identified by PAM clustering. Black dots represent abundance distributions of
bacterial genera from an individual host and numbered white rectangles mark the centre of each enterotype. Bacterial genera that mainly
contribute to each enterotype are listed.
2013). We defined core taxa as taxa present in more
than 90% of our 107 samples, assigned at the lowest
possible taxonomic level. Despite the dynamic nature
of baboon gut microbiota, we found evidence for some
core taxa: three at the family level (Lachnospiraceae,
Peptostreptococcaceae
and
Veillonellaceae)
and
four at the genus level (Faecalibacterium, Prevotella,
Bifidobacterium and Oscillibacter).
To investigate how core microbiota have changed with
host phylogeny, we attempted to identify core gut microbial members of the 57 mammalian species (89 individuals) used in Ley et al. (2008). Given the large variation in
mammalian genomes, diets and lifestyles, it is not surprising that we did not find any core taxon below the phylum
level that is shared by all mammals. However, when we
limited our scope to primates alone, we found two familylevel (Ruminococcaceae and Lachnospiraceae) and
one genus-level (Prevotella) core taxa. Since these taxa
are present in most primates surveyed, these core taxa
were most likely present in the last common ancestor of
primates, suggesting they might be important in the
codiversification of the gut microbiota and the primate
hosts.
Enterotypes in baboons
Previous studies reported that humans and chimpanzees
harbour compositionally similar gut enterotypes
(Arumugam et al., 2011; Moeller et al., 2012). To test for
the presence of enterotypes in our subjects, we clustered
gut microbiota for the 47 (of 107) samples collected from
sexually mature, adult baboons by applying the partitioning around means clustering method on the Bray–Curtis
dissimilarities calculated using genus level abundances
(Arumugam et al., 2011). Our analysis revealed an
optimum of two clusters (Fig. 5, CH index: 39; average
silhouette coefficient: 0.265; prediction strength: 0.79).
Although the silhouette coefficient is comparable to those
reported in earlier enterotype studies (Arumugam et al.,
2011; Moeller et al., 2012), it would be considered low
according to the thresholds proposed more recently
(Koren et al., 2013). Therefore, the enterotypes identified
in this study are tentative. The genera that contributed
most significantly to each cluster were Bifidobacterium,
Butyrivibrio, Megasphaera and Olsenella in enterotype 1
(n = 9) and Oscillibacter and Ruminococcus in enterotype
2 (n = 38). The relative abundances of genera in the adult
© 2015 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology
Gut microbiota in wild baboons
9
Fig. 6. Baboon enterotypes switched over
time. Samples were collected from 1994 to
1999, and in 2001 and 2009, eight (of 14)
adults only had one sample and therefore
were not shown here. Filled rectangles:
enterotype 1; unfilled rectangles: enterotype
2; rectangles with dashed line: sample
missing. When an individual switched
enterotypes during the middle of the year, it is
represented by a hybrid rectangle (half filled
and half unfilled).
samples are listed in Table S7. These two enterotypes
differ from those of humans and chimpanzees. One parsimonious explanation is that enterotypes in humans and
chimpanzees may have evolved since the split between
apes and old world monkeys ∼30 million years ago.
Previous studies found that enterotypes can be
replaced within 1 year in chimpanzees (Moeller et al.,
2012) and within 1 week in wild mice housed in captivity
(Wang et al., 2014). We observed enterotype replacements for most baboons when we assessed the samples
of the same individual at multiple time points (Fig. 6).
Enterotypes changed rapidly in baboons, sometimes
switching in as little as 45 days. Past studies have suggested that proportion of protein versus carbohydrates in
host diet is linked to the host’s enterotype (Wu et al., 2011;
Wang et al., 2014). However, the baboon enterotypes
that we found were not significantly associated with any
factors tested, including diet diversity (richness, Shannon’s H and PCoA axis), age, rainfall, host identity,
season, sex or social group (Wilcoxon rank-sum test
or Fisher’s exact test). Consistent with the finding of
Wang et al. (2014), we found no enterotypes at the OTU
level.
Experimental procedures
Study subjects and predictors of microbiota structure
Study subjects were wild baboons living in the Amboseli
Ecosystem in Kenya, a semi-arid savannah located northeast
of Mt. Kilimanjaro (2°40′S, 37°15′E, 1100 m altitude). Since
1971, the baboons in this area have been studied by the
ABRP (Alberts and Altmann, 2012). Several types of data are
collected throughout the year on known individuals by fulltime, experienced observers, two to three times per week per
group. Here we describe the data collection on the specific
predictor variables we tested; sample sizes vary somewhat
for each predictor variable because some data were only
available for or relevant to some individuals and samples.
Sex, age and developmental stage. Baboons are sexually
dimorphic, and sex is known from conspicuous external
genital morphology. Ages were known to within a few days for
20 of 24 animals in our main dataset (n = 107). The remaining
four individuals immigrated into the population after birth and
their ages were estimated using well-defined metrics and
comparison to known-age animals (Alberts and Altmann,
1995). These four animals had birth dates estimated to be
accurate within 1 year (n = 1), 2 years (n = 2) or 3 years
(n = 1). As baboons mature, they pass through several developmental stages that may also influence the gut microbiota,
including: (i) infancy, during which diet includes both milk and
foods from the environment (from birth to 1.5 years; n = 22
fecal samples from nine individuals), (ii) the juvenile period,
which begins post weaning (∼1.5 years) and ends at sexual
maturity (∼4.5 years for females; ∼5.4 years for males; 38
samples from 10 individuals; Onyango et al., 2013) and (iii)
adulthood, defined by the onset of sexual maturity (n = 47
samples from 14 individuals).
Diet. In addition to the dietary changes associated with the
transition from the infant to the juvenile stage, we tested the
effect of diet composition on gut microbiota. Specifically, for a
subset of subjects (n = 76 fecal samples from 10 juveniles
and 14 adults), we estimated diet composition using behavioural sampling on all the juvenile and adult female members
of the social group in the 30 days prior to sample collection.
Social group members consume similar foods in roughly
similar proportions; hence, group-level diets provide suitable
estimates of the composition of individual diets. The baboons’
diets included 11 food categories: (i) grass, including corms,
blades and grass seed heads, (ii) gum from the bark of
Acacia xanthophloea, (iii) leaves from herbaceous plants or
© 2015 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology
10
T. Ren et al.
trees, (iv) fruits, (v) blossoms, (vi) bark from A. xanthophloea,
(vii) fresh, green seed pods from Acacia spp., (viii) dried
seeds from Acacia spp., (ix) invertebrates, (x) liquid from or
items in or under dung and (xi) unknown unidentifiable items
(Table S3).
We used these data to characterize diet alpha and beta
diversity, noting that time spent feeding is not always proportional to the amount of food ingested. Diet alpha diversity was
measured as both the total number of foods (diet richness)
and dietary Shannon’s H using the vegan package in R
(Oksanen et al., 2013). Diet beta diversity was estimated via
PCoA on a Bray–Curtis dissimilarity matrix of diet composition using vegan. The first three axes of the PCoA explained
80% of the variation in the diet (PC1 = 46%, PC2 = 23%,
PC3 = 11%); PC1 was associated with a tradeoff in relative
proportions of grass (−) versus fruit (+); PC2 was associated
with the proportion of invertebrates (−) versus fruit (+); and
PC3 was associated with the proportion of the diet attributed
to the ‘unknown’ category (−) (Figs S3–S5).
Rainfall. Semi-arid savannah ecosystems are characterized
by highly seasonal patterns of rain that may affect diet as well
as bacterial exposures through sources of drinking water.
Each year, Amboseli experiences a 5 month dry season
(June–October) during which no rain falls. In the remaining 7
months (November–May), the ecosystem receives highly
variable amounts of rain (yearly average = 350 mm;
range = 141–757 mm) (Alberts et al., 2005). The effects of
rainfall were assessed by summing the total amount of rain
that fell in the 30 days prior to sample collection.
Social relationships. One prior study has linked aspects of
primate social group membership to microbial composition
(Degnan et al., 2012). We tested three aspects of social
group: (i) the identity of the animal’s social group on the day
of sample collection, (ii) the size of the animal’s social group
on the day of sample collection, as the number of members
and (iii) the identity of the animal’s natal social group, if
known (in baboons, males are the dispersing sex and the
current group of an adult male invariably differs from his natal
group).
In addition, in baboons, dominance rank has been linked to
physiology and health (Sapolsky and Altmann, 1991; Alberts
et al., 1992; Gesquiere et al., 2011); hence, we also tested for
associations between microbial composition and dominance
rank. Rank was assigned monthly by observing dyadic
agonistic interactions and assigning winners and losers
based on the outcome. These wins and losses were used to
construct dominance matrices, resulting in an ordinal rank for
each member of the group (Hausfater, 1975).
Adult female reproductive status. Prior research has shown
that reproductive cycle changes in human women can influence gut microbial composition (Koren et al., 2012). To test
this idea, samples from adult females (24 samples from nine
individuals) were assigned to one of three reproductive states
(ovarian cycling, pregnant or lactating) using previously published and well-defined criteria (Altmann, 1973; Wildt et al.,
1977; Shaikh et al., 1982; Beehner et al., 2006; Gesquiere
et al., 2007).
Sample collection, DNA extraction and 16S rRNA
sequencing
Gut microbiota were characterized from fecal samples.
Samples for this analysis spanned from 1994 to 2009 and
included 144 samples from 32 individuals. Samples were
chosen to provide both cross-sectional and longitudinal information, including multiple samples from a subset of 13 individuals. All fecal samples were collected within a few minutes
of defecation, after which the sample was mixed and
preserved in 95% ethanol. Samples were stored in an evaporative cooling structure (approximate daily maximum temperature of 25°C) until shipment to the US, where they were
stored at −80°C. DNA was extracted from each sample by
bead beating and phenol-chloroform extraction. For each
DNA extract, the V1–V3 hypervariable regions of the 16S
rRNA gene were PCR amplified and pyrosequenced as
described previously (Ren et al., 2013) on a 454 Life Science
Genome Sequencer FLX platform (University of Virginia
Department of Biology Genome Core Facility).
Sequence processing, quality control and
OTU classification
Sequencing reads were processed using the QIIME pipeline
(Caporaso et al., 2010). Each read was assigned to a sample
by a bar code and then filtered to remove reads with: (i)
lengths less than 200 base pairs or greater than 550 base
pairs, (ii) average Phred equivalent quality scores less than
25, (iii) improper primer or bar code sequences or (iv) the
presence of ambiguous base calls. Eukaryotic, mitochondrial
sequences were removed by BLAST search against the
SILVA database (Quast et al., 2013). Chloroplast sequences
were removed using the Ribosomal Database Project (RDP)
classifier (Wang et al., 2007). Chimeric sequences were identified using UCHIME with the de novo detection algorithm
and default parameters (Edgar et al., 2011). Read filtering
removed 17.8% of the total reads, with the majority removed
as chimeras. The remaining reads were clustered to 97%
OTUs by cdhit (Fu et al., 2012). To further remove potential
sequencing artefacts, we excluded any OTU with ≤5 reads
across all samples. The most abundant sequence of each
OTU was chosen as the representative sequence and classified using the RDP classifier.
Statistical analyses
Of our initial set of 144 samples, three were excluded as
outliers at two or more of four measures of OTU alpha diversity (these outliers had alpha diversity values more than three
times the interquartile range below the lower 25% percentile).
Three additional samples were removed as outliers during
initial beta diversity analyses. During rarefaction, 31 and 84
samples were removed due to insufficient reads, leaving 107
samples (Table 1) and 54 samples (Table S1) for 1500- and
3000-read datasets respectively. Up to 1500- and 3000-read
datasets were compared using the Pearson correlation coefficient and Mantel tests implemented in QIIME pipeline. The
two datasets were found to be highly similar (see Results and
discussion); hence, in the main text we present the results of
the 1500-read dataset.
© 2015 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology
Gut microbiota in wild baboons
Comparison to other mammals. To understand how gut
microbiota from the Amboseli baboons are compared with
other primates and mammals, we conducted PCoA on
unweighted UniFrac matrix to compare one randomly
selected sample from each adult baboon (n = 14) to 89 individual mammals of 57 species surveyed by Ley et al. (2008).
Only the V1–V3 regions of the 16S rRNA gene were compared. Samples were rarefied to 140 reads due to the small
number of reads in the mammal dataset (Ley et al., 2008).
Testing predictors of gut microbial alpha diversity. To test
which factors best predicted microbial alpha diversity, we
constructed linear mixed models of four measures of OTU
alpha diversity: OTU richness (i.e. the number of distinct
OTUs in a sample), Shannon’s H, Chao1 (log transformed)
and Faith’s phylogenetic diversity. All models included host
identity as a random factor; the best-fitting models were identified using the log likelihood criterion.
Testing predictors of gut microbial beta diversity. To investigate the predictors of gut microbial composition, we first
performed exploratory PCoA, followed by hypothesis testing
via CCA (Palmer, 1993). PCoA was performed on unweighted
and weighted UniFrac dissimilarities calculated from the relative abundance of OTUs in each sample (Lozupone and
Knight, 2005). CCA was performed on the relative abundance
of taxa at the phylum, genus, and OTU level and host associated metadata using the vegan package in R (Table 2 and
Table S4). For each test, the best model was selected using
the log likelihood criterion, and the significance of each predictor was assessed by permutation tests. We did not correct
for multiple comparisons in our CCAs because of the nested
nature of these analyses. It would be overly conservative to
account for multiple comparisons because tests of many of the
factors (e.g. age, rainfall) are not independent across models.
To test which factors predicted the relative abundance of the
four most common bacteria phyla, we constructed generalized
linear mixed models with host identity as a random factor, and
a Poisson-distributed error structure. The best-fitting models
were chosen using the log likelihood criterion.
Core microbiota. Since closely related bacterial taxa are
sometimes ecologically interchangeable (Harvey and Pagel,
1991), it may be useful to consider phylogenetic relationships
when identifying core taxa. Core OTUs were identified using
a tree-based algorithm and were defined as those OTUs that
belonged to the same lineage and occurred in more than 90%
of the samples. We identified core OTUs in both baboon and
other mammalian (Ley et al., 2008) gut microbiota.
Enterotype analyses. We performed the enterotype analysis
of the adult baboon gut microbiota as described in Arumugam
et al. (2011), which used Calinski–Harabasz index as an
indicator of optimal clustering. In addition, we calculated the
silhouette coefficient and prediction strength using the
methods suggested by Koren et al. (2013) in R (packages:
cluster, clusterSim, fpc). Genera that mainly contribute to each
enterotype were identified with Randomforest implemented in
QIIME. We consider a genus as a main contributor if its
removal increases >15% overall estimated generalization
error (an estimate of how much error the classifier would have
11
on a novel dataset). Associations between enterotype and
age, rainfall or diet PC axis were tested with Wilcoxon ranksum test. Association between enterotype and host identity,
season, sex or social group was tested with Fisher’s exact test.
Acknowledgements
We would like to thank Zhang Wang and Adam Labonte for
help with 16S PCR amplification and Alex Koeppel for initial
16S rRNA data analysis. We are grateful to Jeanne Altmann,
who has been a director of ABRP since 1971. ABRP is
supported by the National Science Foundation and the
National Institute on Aging. In the last decade, we acknowledge the support from IOS 1053461, IBN 9985910, IBN
0322613, IBN 0322781, BCS 0323553, BCS 0323596, DEB
0846286, DEB 0846532, IOS 0919200, R01AG034513-01
and P01AG031719. We also thank Duke University, Princeton University, the Princeton Center for the Demography of
Aging, and the Max Planck Institute for Demography. We
thank the Kenya Wildlife Services, the Institute of Primate
Research, the National Museums of Kenya, the National
Council for Science and Technology, and members of the
Amboseli-Longido pastoralist communities for their assistance in Kenya. A number of people contributed to the longterm data collection over the years. Particular thanks go to
ABRP’s long-term field team (R.S. Mututua, S. Sayialel and
J.K. Warutere) and to T. Wango for his assistance in Nairobi.
Karl Pinc provided expertise in database design and management. We also thank our database technicians, particularly D. Onderdonk, C. Markham, T. Fenn, N. Learn, L.
Maryott, P. Onyango and J. Gordon. The authors declare not
to have any financial or non-financial competing interests.
References
Ahn, J., Sinha, R., Pei, Z., Dominianni, C., Wu, J., Shi, J.,
et al. (2013) Human gut microbiome and risk of colorectal
cancer. J Natl Cancer Inst 105: 1907–1911.
Alberts, S.C., and Altmann, J. (1995) Balancing costs and
opportunities: dispersal in male baboons. Am Nat 145:
279–306.
Alberts, S.C., and Altmann, J. (2012) The Amboseli Baboon
Research Project: 40 years of continuity and change. In
Long-term Field Studies of Primates. Kappeler, P., and
Watts, D.P. (eds). Heidelberg, Germany: Springer Verlag,
pp. 261–287.
Alberts, S.C., Sapolsky, R.M., and Altmann, J. (1992)
Behavioral, endocrine, and immunological correlates of
immigration by an aggressive male into a natural primate
group. Horm Behav 26: 167–178.
Alberts, S.C., Hollister-Smith, J.A., Mututua, R.S., Sayialel,
S.N., Muruthi, P.M., Warutere, J.K., and Altmann, J. (2005)
Seasonality and long term change in a savannah environment. In Primate Seasonality: Implications for Human Evolution. Brockman, D.K., and Van Schaik, C.P. (eds). New
York, USA: Cambridge University Press, pp. 157–195.
Altmann, S.A. (1973) The pregnancy sign in savannah
baboons. J Zoo Anim Med 4: 8–12.
Amato, K.R. (2013) Co-evolution in context: the importance
of studying gut microbiomes in wild animals. Microbiome
Sci Med 1: 10–29.
© 2015 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology
12
T. Ren et al.
Arumugam, M., Raes, J., Pelletier, E., Le Paslier, D.,
Yamada, T., Mende, D.R., et al. (2011) Enterotypes of the
human gut microbiome. Nature 473: 174–180.
Beehner, J.C., Nguyen, N., Wango, E.O., Alberts, S.C., and
Altmann, J. (2006) The endocrinology of pregnancy
and fetal loss in wild baboons. Horm Behav 49: 688–
699.
Bengmark, S. (2013) Gut microbiota, immune development
and function. Pharmaco Res 69: 87–113.
Bolnick, D.I., Snowberg, L.K., Hirsch, P.E., Lauber, C.L.,
Knight, R., Caporaso, J.G., and Svanback, R. (2014) Individuals’ diet diversity influences gut microbial diversity in
two freshwater fish (threespine stickleback and Eurasian
perch). Ecol Lett 17: 979–987.
Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K.,
Bushman, F.D., Costello, E.K., et al. (2010) QIIME allows
analysis of high-throughput community sequencing data.
Nat Methods 7: 335–336.
Caporaso, J.G., Lauber, C.L., Walters, W.A., Berg-Lyons, D.,
Lozupone, C.A., Turnbaugh, P.J., et al. (2011) Global patterns of 16S rRNA diversity at a depth of millions of
sequences per sample. Proc Natl Acad of Sci U S A 108:
4516–4522.
Codron, D., Lee-Thorp, J., Sponheimer, M., Ruiter, D., and
Codron, J. (2008) What insights can baboon feeding
ecology provide for early hominin niche differentiation? Int
J Primatol 29: 757–772.
David, L.A., Maurice, C.F., Carmody, R.N., Gootenberg, D.B.,
Button, J.E., Wolfe, B.E., et al. (2014) Diet rapidly and
reproducibly alters the human gut microbiome. Nature 505:
559–563.
Degnan, P.H., Pusey, A.E., Lonsdorf, E.V., Goodall, J.,
Wroblewski, E.E., Wilson, M.L., et al. (2012) Factors associated with the diversification of the gut microbial communities within chimpanzees from Gombe National Park. Proc
Natl Acad Sci USA 109: 13034–13039.
Delsuc, F., Metcalf, J.L., Wegener Parfrey, L., Song, S.J.,
Gonzalez, A., and Knight, R. (2014) Convergence of gut
microbiomes in myrmecophagous mammals. Mol Ecol 23:
1301–1317.
DeVore, I., and Washburn, S. (1963) Baboon ecology and
human evolution. In African Ecology and Human Evolution.
Howell, F.C., and Bourliere, F. (eds). Chicago, IL, USA:
Aldine, pp. 335–367.
Dillon, R.J., Vennard, C.T., Buckling, A., and Charnley, A.K.
(2005) Diversity of locust gut bacteria protects against
pathogen invasion. Ecol Lett 8: 1291–1298.
Edgar, R.C., Haas, B.J., Clemente, J.C., Quince, C., and
Knight, R. (2011) UCHIME improves sensitivity and
speed of chimera detection. Bioinformatics 27: 2194–
2200.
Fierer, N., Bradford, M.A., and Jackson, R.B. (2007) Toward
an ecological classification of soil bacteria. Ecology 88:
1354–1364.
Flores, R., Shi, J., Gail, M.H., Gajer, P., Ravel, J., and
Goedert, J.J. (2012) Association of fecal microbial diversity
and taxonomy with selected enzymatic functions. PLoS
ONE 7: e39745.
Fu, L., Niu, B., Zhu, Z., Wu, S., and Li, W. (2012) CD-HIT:
accelerated for clustering the next-generation sequencing
data. Bioinformatics 28: 3150–3152.
Fulthorpe, R.R., Roesch, L.F., Riva, A., and Triplett, E.W.
(2008) Distantly sampled soils carry few species in
common. ISME J 2: 901–910.
Gesquiere, L.R., Wango, E.O., Alberts, S.C., and Altmann, J.
(2007) Mechanisms of sexual selection: sexual swellings
and estrogen concentrations as fertility indicators and cues
for male consort decisions in wild baboons. Horm Behav
51: 114–125.
Gesquiere, L.R., Learn, N.H., Simao, M.C.M., Onyango, P.O.,
Alberts, S.C., and Altmann, J. (2011) Life at the top: rank
and stress in wild male baboons. Science 333: 357–360.
Greenblum, S., Turnbaugh, P.J., and Borenstein, E. (2011)
Metagenomic systems biology of the human gut
microbiome reveals topological shifts associated with
obesity and inflammatory bowel disease. Proc Natl Acad
Sci USA 109: 594–599.
Harvey, P.H., and Pagel, M.D. (1991) The Comparative
Method in Evolutionary Biology. Oxford, UK: Oxford University Press.
Hausfater, G. (1975) Dominance and Reproduction in
Baboons: A Quantitative Analysis. Basel, Switzerland:
S Karger.
Hong, P.Y., Wheeler, E., Cann, I.K., and Mackie, R.I. (2011)
Phylogenetic analysis of the fecal microbial community in
herbivorous land and marine iguanas of the Galapagos
Islands using 16S rRNA-based pyrosequencing. ISME J 5:
1461–1470.
Hooper, L.V., Littman, D.R., and Macpherson, A.J. (2012)
Interactions between the microbiota and the immune
system. Science 336: 1268–1273.
Iida, N., Dzutsev, A., Stewart, C.A., Smith, L., Bouladoux, N.,
Weingarten, R.A., et al. (2013) Commensal bacteria
control cancer response to therapy by modulating the
tumor microenvironment. Science 342: 967–970.
Karlsson, F.H., Tremaroli, V., Nookaew, I., Bergstrom, G.,
Behre, C.J., Fagerberg, B., et al. (2013) Gut metagenome
in European women with normal, impaired and diabetic
glucose control. Nature 498: 99–103.
Koeppel, A.F., and Wu, M. (2012) Lineage-dependent ecological coherence in bacteria. FEMS Microbiol Ecol 81:
574–582.
Koeth, R.A., Wang, Z., Levison, B.S., Buffa, J.A., Org, E.,
Sheehy, B.T., et al. (2013) Intestinal microbiota metabolism
of l-carnitine, a nutrient in red meat, promotes atherosclerosis. Nat Med 19: 576–585.
Koren, O., Goodrich, J.K., Cullender, T.C., Spor, A., Laitinen,
K., Kling Bäckhed, H., et al. (2012) Host remodeling of the
gut microbiome and metabolic changes during pregnancy.
Cell 150: 470–480.
Koren, O., Knights, D., Gonzalez, A., Waldron, L., Segata, N.,
Knight, R., et al. (2013) A guide to enterotypes across the
human body: meta-analysis of microbial community structures in human microbiome datasets. PLoS Comput Biol 9:
e1002863.
Ley, R.E., Hamady, M., Lozupone, C., Turnbaugh, P.J.,
Ramey, R.R., Bircher, J.S., et al. (2008) Evolution of
mammals and their gut microbes. Science 320: 1647–
1651.
Lozupone, C., and Knight, R. (2005) UniFrac: a new
phylogenetic method for comparing microbial communities.
Appl Environ Microbiol 71: 8228–8235.
© 2015 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology
Gut microbiota in wild baboons
Lozupone, C.A., and Knight, R. (2007) Global patterns in
bacterial diversity. Proc Natl Acad Sci USA 104: 11436–
11440.
McKenna, P., Hoffmann, C., Minkah, N., Aye, P.P., Lackner,
A., Liu, Z., et al. (2008) The macaque gut microbiome in
health, lentiviral infection, and chronic enterocolitis. PLoS
Pathog 4: e20.
Markle, J.G.M., Frank, D.N., Mortin-Toth, S., Robertson, C.E.,
Feazel, L.M., Rolle-Kampczyk, U., et al. (2013) Sex
differences in the gut microbiome drive hormonedependent regulation of autoimmunity. Science 339:
1084–1088.
Martínez, I., Muller, C.E., and Walter, J. (2013) Long-term
temporal analysis of the human fecal microbiota revealed a
stable core of dominant bacterial species. PLoS ONE 8:
e69621.
von Mering, C., Hugenholtz, P., Raes, J., Tringe, S.G.,
Doerks, T., Jensen, L.J., et al. (2007) Quantitative
phylogenetic assessment of microbial communities in
diverse environments. Science 315: 1126–1130.
Moeller, A.H., Degnan, P.H., Pusey, A.E., Wilson, M.L., Hahn,
B.H., and Ochman, H. (2012) Chimpanzees and humans
harbour compositionally similar gut enterotypes. Nat
Commun 3: 1179. doi: 10.1038/ncomms2159.
Moeller, A.H., Li, Y., Mpoudi Ngole, E., Ahuka-Mundeke, S.,
Lonsdorf, E.V., Pusey, A.E., et al. (2014) Rapid changes in
the gut microbiome during human evolution. Proc Natl
Acad Sci USA 111: 16431–16435.
Morgan, X.C., Segata, N., and Huttenhower, C. (2013)
Biodiversity and functional genomics in the human
microbiome. TIG 29: 51–58.
Nguyen, N., Gesquiere, L., Alberts, S.C., and Altmann, J.
(2012) Sex differences in the mother–neonate relationship
in wild baboons: social, experiential and hormonal correlates. Anim Behav 83: 891–903.
Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin,
P.R., O’Hara, R.B., et al. (2013) vegan: Community
ecology package. R package version 2.0-7. [WWW
document]. URL http://www.CRAN.R-project.org/package
=vegan.
Onyango, P.O., Gesquiere, L.R., Altmann, J., and Alberts,
S.C. (2013) Puberty and dispersal in a wild primate population. Horm Behav 64: 240–249.
Palmer, M.W. (1993) Putting things in even better order: the
advantages of canonical correspondence analysis.
Ecology 74: 2215–2230.
Philippot, L., Andersson, S.G., Battin, T.J., Prosser, J.I.,
Schimel, J.P., Whitman, W.B., and Hallin, S. (2010) The
ecological coherence of high bacterial taxonomic ranks.
Nat Rev Microbiol 8: 523–529.
Pointing, S.B., Chan, Y., Lacap, D.C., Lau, M.C., Jurgens,
J.A., and Farrell, R.L. (2009) Highly specialized microbial
diversity in hyper-arid polar desert. Proc Natl Acad Sci USA
106: 19964–19969.
Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T.,
Yarza, P., et al. (2013) The SILVA ribosomal RNA
gene database project: improved data processing
and web-based tools. Nucleic Acids Res 41: D590–
D596.
Ren, T., Glatt, D.U., Nguyen, T.N., Allen, E.K., Early, S.V.,
Sale, M., et al. (2013) 16S rRNA survey revealed
13
complex bacterial communities and evidence of bacterial
interference on human adenoids. Environ Microbiol 15:
535–547.
Sapolsky, R.M., and Altmann, J. (1991) Incidence of
hypercortisolism and dexamethasone resistance increases
with age among wild baboons. Biol Psychiatry 30: 1008–
1016.
Shaikh, A., Shaikh, S., Celaya, C., and Gomez, I. (1982)
Temporal relationship of hormonal peaks to ovulation and
sex skin deturgescence in the baboon. Primates 23: 444–
452.
Sponheimer, M., Alemseged, Z., Cerling, T.E., Grine, F.E.,
Kimbel, W.H., Leakey, M.G., et al. (2013) Isotopic evidence
of early hominin diets. Proc Natl Acad Sci USA 110:
10513–10518.
Tap, J., Mondot, S., Levenez, F., Pelletier, E., Caron, C.,
Furet, J.P., et al. (2009) Towards the human intestinal
microbiota phylogenetic core. Environ Microbiol 11: 2574–
2584.
Tung, J., Barriero, L.B., Burns, M.B., Grenier, J.C., Lynch, J.,
Grieneisen, L.E., et al. (in press) Social networks predict
gut microbiome composition in wild baboons. eLife.
Turnbaugh, P.J., Ley, R.E., Mahowald, M.A., Magrini, V.,
Mardis, E.R., and Gordon, J.I. (2006) An obesityassociated gut microbiome with increased capacity for
energy harvest. Nature 444: 1027–1131.
Turnbaugh, P.J., Hamady, M., Yatsunenko, T., Cantarel, B.L.,
Duncan, A., Ley, R.E., et al. (2009) A core gut microbiome
in obese and lean twins. Nature 457: 480–484.
Turroni, F., Peano, C., Pass, D.A., Foroni, E., Severgnini, M.,
Claesson, M.J., et al. (2012) Diversity of bifidobacteria
within the infant gut microbiota. PLoS ONE 7: e36957.
Viaud, S., Saccheri, F., Mignot, G., Yamazaki, T., Daillère, R.,
Hannani, D., et al. (2013) The intestinal microbiota modulates the anticancer immune effects of cyclophosphamide.
Science 342: 971–976.
Wang, J., Linnenbrink, M., Kunzel, S., Fernandes, R.,
Nadeau, M.J., Rosenstiel, P., and Baines, J.F. (2014)
Dietary history contributes to enterotype-like clustering
and functional metagenomic content in the intestinal
microbiome of wild mice. Proc Natl Acad Sci USA 111:
E2703–E2710.
Wang, Q., Garrity, G.M., Tiedje, J.M., and Cole, J.R. (2007)
Naïve Bayesian Classifier for Rapid Assignment of rRNA
Sequences into the New Bacterial Taxonomy. Appl Environ
Microbiol 73: 5261–5267.
Wildt, D.E., Doyle, L.L., Stone, S.C., and Harrison, R. (1977)
Correlation of perineal swelling with serum ovarian
hormone levels, vaginal cytology, and ovarian follicular
development during the baboon reproductive cycle. Primates 18: 261–270.
Wu, G.D., Chen, J., Hoffmann, C., Bittinger, K., Chen, Y.Y.,
Keilbaugh, S.A., et al. (2011) Linking long-term dietary
patterns with gut microbial enterotypes. Science 334: 105–
108.
Yatsunenko, T., Rey, F.E., Manary, M.J., Trehan, I.,
Dominguez-Bello, M.G., Contreras, M., et al. (2012)
Human gut microbiome viewed across age and geography.
Nature 486: 222–227.
Yildirim, S., Yeoman, C.J., Sipos, M., Torralba, M., Wilson,
B.A., Goldberg, T.L., et al. (2010) Characterization of the
© 2015 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology
14
T. Ren et al.
fecal microbiome from non-human wild primates reveals
species specific microbial communities. PLoS ONE 5:
e13963.
Supporting information
Additional Supporting Information may be found in the online
version of this article at the publisher’s web-site:
Fig. S1. The distribution patterns of the unclassified and
classified OTU at the phylum level across the samples.
OTUs have been sorted into bins based on their prevalence
in the samples (X-axis). Y-axis is the count of OTUs in each
bin.
Fig. S2. PCoA analysis of the weighted UniFrac dissimilarities comparing baboon gut microbiota. Each point corresponds to a sample coloured by (A) individual identity, (B)
sex, (C) age class and (D) season, (E) diet group. Baboons
with diet composition information (n = 76) were divided into
three diet groups by the relative abundance of grass, fruit and
invertebrate in their diet guided by the PCoA plot of diet
Bray–Curtis dissimilarity: 1. Fruit, if fruit percentage is >=20%;
2. Invertebrate, if there is invertebrate in diet; 3. Grass, if
grass percentage is >=70%.
Fig. S3. The first principal coordinate of variation in diet
composition (diet PC1) as a function of the 11 primary diet
components (Table S3). Blue lines represent lowess regression fits. PC1 explained 46% of the variation in diet composition and is associated with a tradeoff in the proportion of
grass (−) versus fruit (+) in the baboons’ diets.
Fig. S4. The second principal coordinate of variation in
diet composition (diet PC2) as a function of the 11 primary
diet components (Table S3). Blue lines represent lowess
regression fits. PC1 explained 23% of the variation in diet
composition and is associated with a tradeoff in proportion of
insects (−) versus fruit (+) in the baboons’ diets.
Fig. S5. The third principal coordinate of variation in diet
composition (diet PC3) as a function of the 11 primary diet
components (Table S3). Blue lines represent lowess regression fits. PC1 explained 11% of the variation in diet composition and is associated with the proportion of the diet
attributed to ‘unknown’ categories (−).
Table S1. Sample size information, including the number of
individuals and fecal samples used in analyses of the dataset
rarefied to 3000 reads.
Table S2. Unweighted UniFrac dissimilarity comparison
within and between mammalian orders or diet types.
Table S3. Diet items included in each diet category.
Table S4. CCA analysis of environment and host factors for
the 3000-read dataset.
Table S5. Best-supported generalized linear mixed model
(Poisson link) explaining variation in abundance of the four
most common bacteria phyla for the subset of 76 samples with
diet data. Individual identity is a random effect in all models.
Table S6. Mantel test of correlation between sampling time
interval and microbiota weighted UniFrac dissimilarity
between samples that were collected from the same
individual.
Table S7. Genus level abundance table of 47 adult baboon
samples used in the enterotype analysis.
© 2015 Society for Applied Microbiology and John Wiley & Sons Ltd, Environmental Microbiology